Formulación de un estimador de estados distribuido para el sistema eléctrico usando mediciones sincrofasoriales en nodos referencia.

Abstract

This paper presents the formulation of a distributed state estimator for the prediction of variables in electrical power systems using the weighted least squares method. The estimator uses nodal voltage and angle measurements using phasor measurement units and branch power flow measurements. The development of this proposal was born based on the rapid expansion of electrical networks with their distributed dynamic components and the limitations of a conventional centralized estimation versus the requirements of distributed control techniques. The proposed estimator uses a minimum number of measurements for its correct operation, so the formation of a minimum spanning tree is proposed to obtain the corresponding power flow measurements by areas that guarantee observability in them. Once the mathematical formulation for the distributed state estimator has been developed, it is implemented through the development of a software with the programming tools present in Matlab, which makes use of data in Excel type files (.xlsx extension) from the PowerFactory software, where the test model is implemented. Finally, the results within this work are as expected when compared against a centralized state estimation and the estimation values obtained within PowerFactory, demonstrating that the prediction of electrical variables using the proposed distributed state estimator is highly reliable.